HIV prevention efforts for men who have sex with men (MSM) seem stalled: 18,000 MSM are still diagnosed with Human Immunodeficiency Virus (HIV) annually. MSM under age 25 are especially likely to have unprotected anal intercourse (UAI) with casual partners. Traditional interventions have reduced UAI but these more conscious, deliberative, and cognitive approaches don't address a more automatic, affect-based route to decision-making. Experience with risk cues is needed to produce such automatic, affect-based risk reduction: But, real-life experience could be catastrophic. SOLVE (Socially Optimized Learning in Virtual Environments) is a new approach to HIV prevention that integrates traditional cognitive approaches (e.g., social-cognitive interventions modeling cognitive and behavioral skills), while addressing MSM's affect- based and reactive risky decision-making processes, by giving them experience with risk cues in a safe, virtual environment. This approach has been found to reduce UAI compared to wait-list and "standard of care" one-on-one counseling controls. Across the three ethnic populations of MSM (Black/African-American, Latino/Hispanic, White/Caucasian) in our ongoing work, preliminary findings are stronger for younger (18-24 year old) MSM who take greater risks (i.e., 2 or more UAI with non-primary sex partners in the last 90 days). Furthermore, virtual risk taking was uniquely predictive of future risk-taking, even accounting for traditional self-report measures (e.g., intent, self- efficacy). However, our work using interactive video (SOLVE-IAV) is limited by the number of potential learning situations that IAV technology affords and by IAV's inability to enable a more personalized virtual experience for the user that may enhance the user's sense of "presence" in the experience -- a factor also related to change in UAI. Our first specific aim in the current proposal, following formative research, is to create SOLVE-IT, a virtual environment covering a range of test-situations or "contexts of risk" for diverse MSM that, using Intelligent Agent and Gaming technologies, would be delivered and assessed "on-line" over the web nationally. Our second specific aim is to test the effectiveness of SOLVE-IT for young high risk MSM (Black/African-Americans, Latino/Hispanics, White/Caucasians) compared to a wait-list control group using a 3- month longitudinal randomized control trial (RCT). UAI change with casual partners is the primary dependent variable. Additional exploratory questions are examined. PUBLIC HEALTH RELEVANCE: If this work is successful it would provide further evidence for the effectiveness of an innovative, integrative approach for reducing MSM's sexual risk-taking, that would thereby reduce adverse health outcomes (e.g., HIV transmission). More broadly, the work would advance the science of optimizing personalized risk-reduction, a technology- enabled science that could provide health applications for reducing risky decisions that can adversely impact public health -- all readily available to the public over the web.